Colonization of Northern Eurasia by Modern Humans: Radiocarbon Chronology and Environment
Total Page：16
File Type：pdf, Size：1020Kb
Journal of Archaeological Science (2002) 29, 593–606 doi:10.1006/jasc.2001.0753, available online at http://www.idealibrary.com on Colonization of Northern Eurasia by Modern Humans: Radiocarbon Chronology and Environment P. M. Dolukhanov* Department of Archaeology, University of Newcastle upon Tyne, NE1 7RU, U.K. A. M. Shukurov School of Mathematics and Statistics, University of Newcastle upon Tyne, NE1 7RU, U.K. P. E. Tarasov Department of Geography, Moscow State University, Vorobyevy Gory, Moscow 119899, Russia G. I. Zaitseva Radiocarbon Laboratory, Institute for History of Material Culture, Russian Academy of Sciences, St Petersburg, Dvortsovaya nab. 18, 191186 Russia (Received 30 January 2001, revised manuscript accepted 25 July 2001) The distribution of frequencies of radiocarbon-dated Palaeolithic sites in northern Eurasia shows three peaks of 40–30, 24–18 and 17–1 ka . We argue that these peaks reflect the waves in the colonization of that area by Anatomically Modern Humans stemming from Central and Eastern Europe and caused by environmental stress. 2002 Elsevier Science Ltd. All rights reserved. Keywords: NORTHERN EURASIA, ANATOMICALLY MODERN HUMANS, RADIOCARBON, UPPER PALAEOLITHIC, LAST ICE AGE, PALAEOCLIMATE. Introduction glacial’’), and OIS 3 consisting of the Late Glacial Maximum (LGM) and the Late Glacial Recession. he present article is aimed at the discussion of the initial dispersal of anatomically modern T humans (AMH) in northern Eurasia, as Data and Methods inferred mainly from radiocarbon chronology and Radiocarbon age palaeoenvironmental evidence. In this respect this article may be viewed as a direct development of the The main resource of the present study consists of the previous publication, focused on East European Plain database of radiocarbon measurements of Siberia’s (Dolukhanov, Shukurov & Sokoloff, 2001). The area Palaeolithic sites compiled at the Radiocarbon Labora- considered here extends to Siberia and Russian Far tory of the Institute for History of Material Culture, East, which are viewed against the wider background Russian Academy of Sciences, St Petersburg. It exceeds of the Eurasian continent. The period of time consid- in scope the previously published compendia by ered encompasses the episodes of the Last Ice Age in Lisitsyn & Svezhentsev (1997) and Kuzmı´n & the range between ca 50,000 and 10,000 years before Tankersley (1997). The measurements included in the present (50–10 ka). This includes Oxygen Isotope Stage database have been performed, in most cases, with (OIS) 2 with the ‘‘megainterstadial’’ (or ‘‘interpleni- the conventional technique in the institutions of the Russian Academy of Sciences: at Institute of Geology *For correspondence. E-mail: [email protected]. Tel.: in Moscow (GIN), Institute of Geography in Moscow +44 (0) 191 222 7848; Fax: +44 (0) 191 222 8561. (IGAN), Institute of Ecology and Evolution in 593 0305–4403/02/$-see front matter 2002 Elsevier Science Ltd. All rights reserved. 594 P. M. Dolukhanov et al. Moscow (IM), Institute for History of Material Cul- or autonomous republic/autonomos oblast (Altai, ture in St Petersburg (LE), Laboratory for Quaternary Khakassia, Yakutia); geographical co-ordinates Geology and Geochronology in Magadan (MAG), the (northern latitude X and eastern longitude Y in degrees Vernadsky Institute of Geochemistry in Moscow and minutes); dated material (charcoal, antler or (MO), and Institute of Geology and Mineralogy in bone), and the stratigraphic position. Apparently most Novosibirsk (SOAN). Several important series were of the dated bone and antler do not have any clear measured in Groningen University, Holland (GrA and signs of human modification. GrN). The database includes samples from several Using also the previously published data for East sites measured with the use of Accelertor Mass European Plain (EEP) (Dolukhanov, Shukurov & Spectrometry (AMS) in several laboratories including Sokoloff, 2001), we have presented the radiocarbon the Oxford Radiocarbon Accelerator Unit, U.K. dates in the frequency histograms showing the occur- (OxA), University of Arizona, U.S.A. (AA) and others. rences of the radiocarbon dated sites per 1000 years. The original database has been screened with the use These are shown for EEP in Figure 1, SCS in Figure 2, of procedures discussed by Dolukhanov, Shukurov & and NES in Figure 3. The Figures also show a com- Sokoloff (2001): posite record of oxygen isotope variations from the sediment cores in the Atlantic Ocean (Imbrie et al., (1) All dates where association with archaeological 1990), with larger 18O values indicating colder climate deposits has not been demonstrated were rejected. throughout the Northern Hemisphere. (2) Wherever possible, preference was given to the As demonstrated by Dolukhanov, Shukurov & dates supplied with adequate stratigraphic and/or Sokoloff (2001), calibration of the radiocarbon dates planigraphic information. belonging to the periods discussed (allowing for varia- (3) Equally, we preferred the measurements con- tions in the 14C production rate in the exchange firmed by inter-laboratory cross-checking. reservoir) does not affect the general trends, and so our (4) Controversial dates (e.g., those failing inter- conclusions. Some statistical aspects of our arguments laboratory cross-checking or a single date are discussed in Appendix 1. strongly deviating from a series of close-by dates for the same site) were discarded. Such a screening is necessary because some sites have Palaeoenvironment vastly different number of radiocarbon date determi- Apart from the oxygen isotope data, we discuss evi- nations. For example, 14 dates have been published for dence pertinent to the Late Quaternary climate and the Berelyoh site, one of them being 42·0 ka (with environment of the major areas in Northern Eurasia, insufficient stratigraphic information and published as well as to the immediate setting of the sites. The without its error) and the remaining ones in the range climate reconstructions are based on the existing 13·4–10·3 ka. Our screened date-list includes just one publications where descriptive qualitative and/or date of 11·8 ka (No. 105 in Table 2) that is close to quantitative non-statistical methods were used (e.g. the median of the date cluster of the younger age. Frenzel, Pecsi & Velichko, 1992). In several cases, The geographical areas discussed here (the Altai quantitative characteristics of the climate were based Mountains, the Yenisei, Angara and Aldan Rivers, the on the concept of biomization: pollen spectra grouped Lake Baikal area and the Maritime Region) have been into plant functional types (PFT) were calibrated rather thoroughly surveyed by several generations against the present-day climatic variables (Prentice of Russian archaeologists. The frequencies of the et al., 1992; Peyron et al., 1998; Tarasov et al., 1999a). screened dates per chosen time interval can be used as a reliable indicator of the population (or settlement) density at a given time only provided each settlement European Russia (or rather a distinct stage of its existence) enters the analysis with equal weight, e.g., as a single date per Palaeoenvironment settlement per chronological stage. A few examples of The climate of the Middle Wu¨rm/Valdai ‘‘megainter- the screening are given below. stadial’’ (OIS 3), which lasted in European Russia from As a result, two new data-lists have been developed, 48–25 ka, was cool and unstable with at least five one for Southern-Central Siberia (SCS. 97 dates), and milder oscillations (Arslanov, 1992). According to North-Eastern Siberia and Russian Far East (NES. 22 Frenzel, Pecsi & Velichko (1992), summer and winter dates), shown in Tables 1 and 2. The data-lists include temperatures in Eastern Europe were lower than today the following information: entry number; laboratory by 4–6C and 4–10C, respectively. Annual precipita- index as defined above; site name; date (uncalibrated tion of that period is estimated as 150–250 mm, also ); the date uncertainty; cluster size (the number of lower than now. dates in the original database from which the date The Last Glacial Maximum (LGM), 20–18 ka, shown is chosen using the screening criteria discussed featured the maximum extension of ice sheets on above); location with respect to Russia’s administrative East European Plain. The biomization-based quantita- division, i.e., oblast/kray (e.g., Irkutsk, Krasnoyarsk) tive assessment (Tarasov et al., 1999a) indicates a Colonization of Northern Eurasia by Modern Humans 595 considerable depression of temperatures, by 20–29Cin ‘‘eastern Gravettian’’. Grigor’ev (1993) has identified, winters and 5–10C in summers. The annual precipita- among them, the ‘‘Kostenki-Avdeevo’’ Culture expos- tion resulting from this approach is 200–450 mm less ing stylistic similarities with the sites in Central than today, with the drought index showing particu- Europe, notably, Willendorf, Predmosti and Dolni larly dry conditions in northern and mid-latitude Vestonice. Russia. The vegetation was dominated by the ‘‘peri- glacial’’ tundra and cold-resistant steppe in combina- The later stage. The latest recognizable increase in the tion with open woodland of larch and birch (Grichuk, density of UP sites occurred at 18–15 ka. At that stage 1992). cultural fragmentation became apparent, the cultural The Late Glacial Recession at 18–10 ka was marked entities being restricted to major river basins: the by the degradation of ice sheets, punctuated by short- Prut-Dniestrian, Upper Dnieprian, Donian,